Abstract:
Objective To improve image quality of low dose CT in whole body PET/CT using adaptive statistical iterative reconstruction (ASIR) technology.
Methods Twice CT scans were performed with GE water model, scan parameters were:120 kV, 120 and 300 mA respectively.In addition, 30 subjects treated with PET/CT were selected randomly, whole body PET/CT were performed after 18F-FDG injection of 3.70 MBq/kg, Sharp IR+time of flight+VUE Point HD technology were used for 1.5 min/bed in PET; CT of spiral scan was performed under 120 kV using automatic exposure control technology (30-210 mA, noise index 25).Model and patients whole body CT images were reconstructed with conventional and 40% ASiR methods respectively, and the CT attenuation value and noise index were measured.
Results Research of model and clinical showed that standard deviation of ASiR method in model CT was 33.0% lower than the conventional CT reconstruction method (t=27.76, P < .01), standard deviation of CT in normal tissues (brain, lung, mediastinum, liver and vertebral body) and lesions (brain, lung, mediastinum, liver and vertebral body) reduced by 21.08%(t=23.35, P < 0.01) and 24.43%(t=16.15, P < 0.01) respectively, especially for normal liver tissue and liver lesions, standard deviations of CT were reduced by 51.33%(t=34.21, P < 0.0) and 49.54%(t=15.21, P < 0.01) respectively.
Conclusion ASiR reconstruction method was significantly reduced the noise of low dose CT image and improved the quality of CT image in whole body PET/CT, which seems more suitable for quantitative analysis and clinical applications.